B-Type Natriuretic Peptide Predict Mortality in Diabetic Patients with Foot Ulcer

The background of this article is to retrospectively investigate the role of B-type natriuretic peptide (BNP) levels as an independent prognostic marker for mortality in diabetic patients with foot ulcer. In the period 1 January-30 June 2009, 279 diabetic patients were admitted to our foot center because of foot ulcer: neuroischemic in 182 (65.2%), neuropathic in 97 (34.8%), infected in 233 (83.5%). In all patients BNP (normal value: ≤ 100 pg/mL) was evaluated at admission. Mortality from any cause was established and the relationship between the recorded variables and risk of death was evaluated by the survival Cox regression model. Kaplan-Meier analysis was performed. BNP was >100 pg/mL in 167 (59.9%) patients. 278 (99.6%) patients were followed to December 31, 2012 with a mean follow-up of 3.6 ± 0.2 years. Seventy-one/278 (25.5%) patients died. Of these, 61 (85.9%) had a BNP value >100 pg/mL. Forty-five/71 (63.4%) patients died from cardiac cause. The most frequent cause of death was heart disease: 45/71 (63.4%). Of these, 24 patients had no history of CAD. Multivariate analysis showed the independent predictors of death to be BNP > 100 (hazard ratio 6.040; p <0.001; CI, 2.380-15.328), age (hazard ratio 1.043, p = 0.024, CI 1.001-1.082), and dialysis treatment (hazard ratio 3.145; p = 0.032; CI, 1.100-8.988). Journal of Research in Diabetes 2


Several
observational studies have suggested a high mortality rate in patients with diabetic foot ulceration (DFU) (1,2).Increased risk of mortality is usually ascribed to a high incidence of cardiovascular morbidity in patients with neuroischemic foot (3,4).Although increased mortality in diabetic patients with neuropathic foot ulcer has received some attention, the identification of highrisk patients remains a priority (5,6).Measurement of markers which predict excess mortality might improve the prediction of adverse outcome in this population of patients.B-type natriuretic peptide (BNP) level is an independent marker both of cardiovascular outcome and also of all cause mortality beyond conventional cardiovascular risk factors (7,8).
In 1999 we started to determine the BNP in diabetic patients admitted to our institution because of a foot ulcer as a marker of increased risk of mortality in patients undergoing foot surgery.Having been very low mortality in the early period the determination of BNP for this purpose has been suspended.However, we followed these patients over time.
We aimed to investigate the role of BNP levels as an independent prognostic marker for all-cause mortality in diabetic patients with foot ulcer.

Data Collection
This is a retrospective study of diabetic patients admitted because of foot ulcer in thr Diabetic Foot Centre in Multimedica Hospital, Sesto S. Giovanni, Italy.The data presented derived from the data base of

Protocol
In all diabetic patients referred to our diabetic foot center for foot ulcer the severity of lesion was graded using the University of Texas Wound Classification System (9).All patients were assessed for the presence of diabetic sensory neuropathy and/or critical limb ischemia (CLI).Sensory neuropathy was detected if vibration perception threshold was > 25 V and sensitivity was absent in >5/9 foot points at Semmes-Weinstein 10 g filament.CLI was detected using the TASC II criteria of non invasive parameters: transcutaneous oxygen tension (TcPO2; TCMTM3, Radiometer GMBH, Copenhagen, Denmark) at the dorsum of the foot was < 30 mmHg and ankle pressure < 70 mmHg when measurable (absent foot pulses or non-compressible foot arteries were due to medial calcifications) with a continuous wave Doppler instrument (DIADOP 50, Mediland s.r.l.Varedo, Milan, Italy) (10).All patients with CLI were referred for an angiographic study and peripheral angioplasty (PTA) or bypass graft (BPG) was performed.After revascularization antithrombotic therapy with ticlopidine 500 mg plus acetyl salicic acid 100 mg was prescribed indefinitely or was confirmed if already used.
The foot lesion was treated with surgical intervention and/or dressing (11) For each patient, all of the following variables were recorded: gender and age; duration and treatment of diabetes; history of cardiac disease, stroke, arterial hypertension and diabetic retinopathy.In this study blood glucose level (mg/dl), glycosylated hemoglobin (HbA1c, Hb%, high pressure liquid chromatography, normal value: 4-6.4),serum creatinine value (mg/dl, Jaffe, Roche, Milan, Italy), arterial systolic and diastolic values, Creactive protein (CRP, mg/dL ulti med After hospital discharge all the patients with foot ulcer were examined at our center until complete ulcer healing occurred.Our Institute is a tertiary diabetic foot center: patients were referred to our center from other hospitals and once healed, they were moved to their original center, for the control of diabetes and treatment of comorbid conditions.All data regarding vital status and cause and date of death were obtained from clinicians managing these patients.

Follow-Up
The vital status of 278 followed patients was determined at the end of follow-up and dates and causes of deaths were recorded.After hospital discharge all the patients with foot ulcer were examined at our Centre until complete ulcer healing occurred.Patients were examined once year in our Centre and vital status was recorded in this moment.All the data regarding patients who did not attend at follow-up visit were obtained from the colleagues treating these patients.

StatisticAnalysis
This is a retrospective study.We considered the consecutive series of patients as a cohort study and reported the descriptive statistics as average values and standard deviation for the continuous variables and as percentages and contingency tables for the qualitative and discrete variables.The 95% level was been adopted to build up the confidence intervals and 5% level has been considered to test the null hypotheses.The time and status of each patient from the date of enrolment to the foot ulcer and death was been studied by applying the Kaplan-Meier approach and the product/ limit curve was built up.The relationship between the recorded variables and risk of death was evaluated by the survival Cox regression model and the data were reported by hazard ratio (HR) and 95% confidence interval (CI).Stata 10.0 (Statistics/Data Analysis, Stata Corporation, 4905 Lakeway Drive, College Station, Texas, USA) was used for calculations.

Results
In the period 1 January-

Discussion
Diabetic patients with foot ulcer have an excess risk of all-cause mortality (12).This data is well known for a long time (13), and a recent meta-analysis has confirmed this (14).Measurement of BNP has proved to be a very reliable marker in the diagnosis of heart failure (15) and is recommended in patients with symptoms of left ventricular dysfunction (16).Moreover BNP is a strong predictor of cardiovascular mortality and cardiovascular disease beyond conventional risk factors (17)(18)(19).
Very recently a number of studies have investigated the value of BNP in diabetes disease (20)(21)(22)(23)(24) but role of BNP in the prognosis of diabetic patients with foot ulcer has not been examined except after the use of hyperbaric oxygen therapy (25).This study has shown that patients with BNP > 100 pg/mL mg/dl have a higher risk of all-cause mortality in patients with foot ulcer.An increased risk was found in aged and dialysis patients, and this is wellknown in the literature.
None of the characteristics related to foot care as infection (CRP; leucocytes number), class of Texas wound classification, revascularization procedures, are associated with mortality.Presence of critical limb ischemia is significantly associated with death at univariate analysis, but is not a result independent risk factor at multivariate analysis.
We did not observe an independent association of risk with history of heart disease ad admission, despite a significant association in univariate analysis.This finding may seem at odds with the reported data in the literature and in particular with the only known to us study on mortality risk, which was performed in type 2 diabetic patients (26).This data is not easily interpreted.We must consider that elevated BNP is present in  (27).One limitation of our study was not to have investigated a subclinical left ventricular hypertrophy or ischemic coronary disease (28,29).However, also the above-mentioned study of Tarnow et al (26) did not carry out investigations beyond at resting electrocardiogram which has a notoriously low sensitivity..However, since BPN has a strong prognostic power, an abnormal value of it could be the reason for starting with a more accurate heart diagnostic evaluation.A simple blood sample could identify asymptomatic patients and could justify more expensive cardiologic imaging tests.Actually, this must be demonstrated by a study "ad hoc".
In our study population only one patient died at 30 days and in the first months: therefore the ability of pre-operative BNP value to predict the at 30 days mortality has not been put to the test (30).

Limitations ofthe Study
This is a retrospective study with its limits.In particular, we did not analyze the presence of silent CAD with proper imaging tests.Then, we do not know if patients with no known CAD history had a silent CAD: it would have added these patients to the patients with CAD history.Again, we did not evaluate the presence of a subclinical left ventricular hypertrophy: in these patients BPN is an important marker.

Conclusions
These findings show that BNP is a strong predictor of mortality in diabetic patients with foot ulcer, not related to the etiology and cure of the foot lesion.Elevated BNP is present in many patients who died from cardiac causes but without history of heart disease: this may suggest the usefulness of early diagnosis of heart disease in these patients?It remains to show that this early diagnosis can then decrease mortality in these patients.

Table 3 : UnivariateAnalysis Results of the Association with the Variables Considered and Death (N = 71) in Study Population (N = 278)
Journal of Research in Diabetes _______________ Ezio Faglia, Giacomo Clerici, Maurizio Caminiti, Curci Vincenzo, Alberto Morabito, Vincenzo Prisco, Marco Prisco, Rosaria Greco and Michael Edmonds (2013), Journal of Research in Diabetes, DOI: 10.5171/2013.388970more than half of the patients without history of cardiac disease died from cardiac causes.One cause could be the presence of subclinical atherosclerosis in patients with elevated BNP value